The Entity Linking (EL) task identifies entity mentions in a text corpus and
associates them with a corresponding unambiguous entry in a Knowledge Base. The
evaluation of EL systems relies on the comparison of their results against gold
standards. A common format used to represent gold standard datasets is the NLP
Interchange Format (NIF), which uses RDF as a data model. However, creating
gold standard datasets for EL is a time-consuming and error-prone process. In
this paper we propose a tool called NIFify to help manually generate, curate,
visualize and validate EL annotations; the resulting tool is useful, for
example, in the creation of gold standard datasets. NIFify also serves as a
benchmark tool that enables the assessment of EL results. Using the validation
features of NIFify, we further explore the quality of popular EL gold standards.